Malignant gliomas are often debilitating, rapidly fatal cancers. Few consistent factors influencing glioma etiology or prognosis have emerged, likely due to heterogeneity of causal and progression pathways within and between histologic types. Many studies suggest a role for inherited susceptibility and environmental and immunologic factors in gliomagenesis. Inherited variation and tumor markers may also influence survival, in addition to established factors such as age, radiation therapy, and resection. This proposal builds on 15 years of studies to increase understanding of these and other factors associated with adult glioma etiology and prognosis among well-characterized cases and controls. In this two phase study, we will genotype constitutive DNA from 900 existing astrocytic cases and 700 controls for about 10,000 single nucleotide polymorphisms (SNPs) and confirm important SNPs in a new group of subjects . We will (1) use functional genomic and bioinformatic tools (e.g. GenMAPP, SIFT , HapMap) to interrogate and integrate genomic and function databases to select SNPs based on clearly defined prioritization criteria (validated, non-synonymous coding or other functional SNPs with allele frequency >5% prioritized for candidate chromosomal regions, genes and pathways potentially relevant to glioma pathogenesis with the remaining spanning the genome); (2) using genotyping results from the 10,000 SNPs, identify 400-1500 SNPs potentially associated with case control status (logistic regressions), case survival (Cox regressions), and age at case diagnosis (analyses of variance) using permutation tests to identify subsets of SNPs with low false discovery rates and other modeling methods such as Random Forests, logic regression and genomic tools to define interacting SNPs and enhance true discovery; (3) (a) accrue another 950 (about 500 population based and 450 clinic based) glioma cases diagnosed from 5/06-12/09 and 500 population-based controls obtaining blood, buccal, and tumor specimens, and treatment and questionnaire data about family and personal medical histories, occupation, and dietary preferences, (b) classify tumors with neuropathology review and tumor markers (genetic and expression changes in p53 and EGFR), (c) conduct IgE and virologic IgG serologic studies, and (d) genotype the new astrocytic cases and controls for the previously identified SNPs (and other SNPs that may be in strong linkage disequilibrium with them); and (4) integrate SNP, demographic, environmental, immunologic, treatment and tumor data as appropriate to determine factors related to glioma etiology and prognosis. Public health significance: Primary brain cancers kill about 13,000 Americans a year; they rank first among all cancer sites for a verage years of life lost; little progress has been made in improving glioblastoma survival in over 30 years; this study will help identify factors that cause these cancers and that may predict survival. Enhanced understanding of these factors may provide targets for future interventions. [unreadable]